ASL recognition using YOLOv7

The previous code will be used to launch a notebook and an app from OVHcloud AI Tools.

📄 Access to Weights & Biases report here.

Requirements

• a Public Cloud project in your OVHcloud account
• access to the OVHcloud Control Panel
• ovhai CLI installed
• a Public Cloud user with administrator or AI Training admin role, see here for more information

You will find all information on OVHcloud documentation.

Create Object Storage container

• Create the data container (empty): ovhai data upload gra data-sign-language

• Create the model container (empty): ovhai data upload gra model-sign-language

• Create the images container (with your own images to do detection on your future model): ovhai data upload gra images-sign-language repo-local-my-test-images/ --remove-prefix repo-local-my-test-imagesl/

Launch an AI Notebook

To launch and access to the AI Notebook, you have to launch the following command:

ovhai notebook run miniconda jupyterlab \
	--name notebook-yolov7-asl \
	--framework-version conda-py39-cuda11.2-v22-4 \
	--gpu 1 \
	--volume data-sign-language@GRA/:/workspace/data:RW:cache \
	--volume model-sign-language@GRA/:/workspace/asl-yolov7-model:RW \
	--volume images-sign-language@GRA/:/workspace/images:RO \
	--volume https://github.com/eleapttn/yolov7_streamlit_asl_recognition.git:/workspace/github-repo:RW

Launch an AI Deploy app

To launch and access to the AI Deploy app, you have to launch the following command:

First, you have to build and push your Docker image to your Docker Hub!

ovhai app run <your_docker_id>/yolov7-asl-recognition:latest \
	--gpu 1 \
	--default-http-port 8501 \
	--volume asl-volov7-model@GRA/:/workspace/asl-yolov7-model:RO

References

Access to the resources:

• Slides are available here.
• You can check the replay on this link.